Integer, Intent (In)	::	n, nrhs, lda, ipiv(*), ldb
Integer, Intent (Out)	::	info
Real (Kind=nag_wp), Intent (In)	::	a(lda,*)
Real (Kind=nag_wp), Intent (Inout)	::	b(ldb,*)
Character (1), Intent (In)	::	uplo

C Header Interface

#include nagmk26.h

void	f07mef_ ( const char uplo, const Integer n, const Integer nrhs, const double a[], const Integer lda, const Integer ipiv[], double b[], const Integer ldb, Integer info, const Charlen length_uplo)

The routine may be called by its LAPACK name dsytrs.

3

Description

f07mef (dsytrs) is used to solve a real symmetric indefinite system of linear equations

A X = B

, this routine must be preceded by a call to f07mdf (dsytrf) which computes the Bunch–Kaufman factorization of

A

uplo ='U'

A = P U D U^{T} P^{T}

, where

P

is a permutation matrix,

U

is an upper triangular matrix and

D

is a symmetric block diagonal matrix with

1

1

and

2

2

blocks; the solution

X

is computed by solving

P U D Y = B

and then

U^{T} P^{T} X = Y

uplo ='L'

A = P L D L^{T} P^{T}

, where

L

is a lower triangular matrix; the solution

X

is computed by solving

P L D Y = B

and then

L^{T} P^{T} X = Y

4

References

Golub G H and Van Loan C F (1996) Matrix Computations (3rd Edition) Johns Hopkins University Press, Baltimore

5

Arguments

1: $uplo$ – Character(1)Input

On entry: specifies how

A

has been factorized.

$uplo ='U'$: $A = P U D U^{T} P^{T}$ , where $U$ is upper triangular.
$uplo ='L'$: $A = P L D L^{T} P^{T}$ , where $L$ is lower triangular.

Constraint:

uplo ='U'

'L'

2: $n$ – IntegerInput

On entry:

n

, the order of the matrix

A

Constraint:

n \geq 0

3: $nrhs$ – IntegerInput

On entry:

r

, the number of right-hand sides.

Constraint:

nrhs \geq 0

4: $a (lda *)$ – Real (Kind=nag_wp) arrayInput

Note: the second dimension of the array a must be at least

\max (1, n)

On entry: details of the factorization of

A

, as returned by f07mdf (dsytrf).

5: $lda$ – IntegerInput

On entry: the first dimension of the array a as declared in the (sub)program from which f07mef (dsytrs) is called.

Constraint:

lda \geq \max (1, n)

6: $ipiv (*)$ – Integer arrayInput

Note: the dimension of the array ipiv must be at least

\max (1, n)

On entry: details of the interchanges and the block structure of

D

, as returned by f07mdf (dsytrf).

7: $b (ldb *)$ – Real (Kind=nag_wp) arrayInput/Output

Note: the second dimension of the array b must be at least

\max (1, nrhs)

On entry: the

n

r

right-hand side matrix

B

On exit: the

n

r

solution matrix

X

8: $ldb$ – IntegerInput

On entry: the first dimension of the array b as declared in the (sub)program from which f07mef (dsytrs) is called.

Constraint:

ldb \geq \max (1, n)

9: $info$ – IntegerOutput

On exit:

info = 0

unless the routine detects an error (see Section 6).

6

Error Indicators and Warnings

$info < 0$: If $info = - i$ , argument $i$ had an illegal value. An explanatory message is output, and execution of the program is terminated.

7

Accuracy

For each right-hand side vector

b

, the computed solution

x

is the exact solution of a perturbed system of equations

(A + E) x = b

, where

if $uplo ='U'$ , $|E| \leq c (n) ε P |U| |D| |U^{T}| P^{T}$ ;
if $uplo ='L'$ , $|E| \leq c (n) ε P |L| |D| |L^{T}| P^{T}$ ,

c (n)

is a modest linear function of

n

, and

ε

is the machine precision.

\hat{x}

is the true solution, then the computed solution

x

satisfies a forward error bound of the form

\frac{{‖x - \hat{x}‖}_{\infty}}{{‖x‖}_{\infty}} \leq c (n) cond (A, x) ε

where

cond (A, x) = {‖|A^{- 1}| |A| |x|‖}_{\infty} / {‖x‖}_{\infty} \leq cond (A) = {‖|A^{- 1}| |A|‖}_{\infty} \leq κ_{\infty} (A)

Note that

cond (A, x)

can be much smaller than

cond (A)

Forward and backward error bounds can be computed by calling f07mhf (dsyrfs), and an estimate for

κ_{\infty} (A)

(

= κ_{1} (A)

) can be obtained by calling f07mgf (dsycon).

8

Parallelism and Performance

f07mef (dsytrs) makes calls to BLAS and/or LAPACK routines, which may be threaded within the vendor library used by this implementation. Consult the documentation for the vendor library for further information.

Please consult the X06 Chapter Introduction for information on how to control and interrogate the OpenMP environment used within this routine. Please also consult the Users' Note for your implementation for any additional implementation-specific information.

9

Further Comments

The total number of floating-point operations is approximately

2 n^{2} r

This routine may be followed by a call to f07mhf (dsyrfs) to refine the solution and return an error estimate.

The complex analogues of this routine are f07msf (zhetrs) for Hermitian matrices and f07nsf (zsytrs) for symmetric matrices.

10

Example

This example solves the system of equations

A X = B

, where

A = (\begin{matrix} 2.07 & 3.87 & 4.20 & - 1.15 \\ 3.87 & - 0.21 & 1.87 & 0.63 \\ 4.20 & 1.87 & 1.15 & 2.06 \\ - 1.15 & 0.63 & 2.06 & - 1.81 \end{matrix}) and B = (\begin{matrix} - 9.50 & 27.85 \\ - 8.38 & 9.90 \\ - 6.07 & 19.25 \\ - 0.96 & 3.93 \end{matrix}) .

Here

A

is symmetric indefinite and must first be factorized by f07mdf (dsytrf).

NAG Library Routine Document

f07mef (dsytrs)

▸▿ Contents

1 Purpose

2 Specification

3 Description

4 References

5 Arguments

6 Error Indicators and Warnings

7 Accuracy

8 Parallelism and Performance

9 Further Comments

10 Example

10.1 Program Text

10.2 Program Data

10.3 Program Results

1

Purpose

2

Specification

3

Description

4

References

5

Arguments

6

Error Indicators and Warnings

7

Accuracy

8

Parallelism and Performance

9

Further Comments

10

Example

10.1

Program Text

10.2

Program Data

10.3

Program Results